enwrit - Agent-first tooling that plugs into any IDE or AI CLIAI agents do what you tell them. Tell them better. Lint instructions, review plans, check doc health, and route context across repos. Agent-first tooling that plugs into any IDE or AI CLI.What writ does

Lint instructions -- Score any AI instruction on six quality dimensions (clarity, structure, coverage, brevity, examples, verification). Research shows 97.1% of AI instructions have quality defects. writ finds them.

Review plans -- Catch risks, missing steps, and bad ordering before your agent executes them. Prompt-injection review via your IDE's LLM -- zero API cost.

Check documentation health -- Detect dead references, treeview drift, and stale docs. Keep AGENTS.md and .cursor/rules/ accurate so your agents stay grounded.

Route context across tools -- Write to Cursor, Claude Code, Copilot, Codex, Windsurf, Kiro -- all from one source. Every tool speaks the same agent.

Agent-to-agent communication -- Hand off context between agents across repos and devices. MCP-native, CLI-first.

Works with what you knowwrit reads and writes the files your tools already use: AGENTS.md, .cursor/rules/, CLAUDE.md, copilot-instructions.md. One source, every format.Install: pip install enwritGitHub | Browse instructions | Threads | About